I'd suggest to check some trn files in irfanview for example (load as raw file): Thus you might get more insight how this works.

(guess, the texture format uses some compression)

ChromeHounds has a bunch of archive files of the format BND - sleet01 and MrAdults were able to make a script that extract these in this thread . yeah, 7 years old thread - welcome to the "Riders of death horses" - club! :D and SMD4 not related to Source Engine? not sure; SMD from Valve i...

maybe you're just searching for things (apps) that don't exist or never will come true?

It's always your own efforts that lead you to somewhere, or a point cloud at least:

I temporary inserted the (if (uvpos>0.5) condition) into the code to get the right part of the image.) Oh i see, so I cant rebuild it! Then I need to find a way to build an .obj for that somehow. Any idea how I could achieve that? In hex2obj switch to ShortAll, then File / SaveAs mesh Load the crea...

However, i can not rebuild das, with your tool. with hex2obj you get the left part of the image in my previous post when pressing the 'uvs' button. (I temporary inserted the (if (uvpos>0.5) condition) into the code to get the right part of the image.) Also: How did you find out to subtract 1.0 for ...

Charles01 wrote:I had no idea what to do at the vertex step.

before creating scripts I'd suggest to get some understanding of the npki model's 3D format:
(using hex2obj, view link in my sig)

this looks like uvs; not sure how to unscramble them atm: (well - done; had to subtract 1.0 from every uv value > 0.5)

I was not able to find the offset at the beginning of the file to these blocks, I can not understand where they can be in * .efo ((( you don't need to; just search for 1A 27 04 00 as a signature in the efo file, add 10 bytes to the offset you'll find. Should result in the start of face block size; ...

guess bumpers will lead to nothing :) I erased the superfluous faces: sunGlasses.jpg then saved in somewhat format to compare which face indices were the wrong ones. But blender's obj export mixes everything up, no chance to do a simple comparison. I remember having created a "render to vertex&...

well, just another format that doesn't make sense to me; there's a good chance that it's animation data. For the face indices, seems you got the wrong ones. i'd try the ones from 0x2B58 to 0x3AA0 size of block: 0xF48 -> 3912/2= 1956 face indices count resulting in an assumed vertex count of 421. The...

Fake tri strips used:

Bump. Its been some time, yeah, it surely has. While it's easy to get some decent point cloud (277 vertices) the face indices are a mess. You have to find out how they have to be used. car1_PMD.JPG I tried out triangle strips with some lines commented out (#) seeming unlogical to me but they are pa...

(you'll need a script for getting the face indices)

Oh! Do you think you could explain how it works? as I wrote: skip the sequence. (As you can see from the right part of the pic there's another rule required but I'm too busy atm to search for it.) And so faces are starting at 0x193C? why? I used 0xFDC for the sunglasses, iirc (but not sure, at leas...

but... let's say i remove parts on model, how do i save changes on g1m or bin? that's the hard part :D Generally the "reverse way" requires a FULL format analysing. If it's removing only (not changing) there's a chance to set the concerning vertices to 0.0 0.0 0.0 or set the face indices ...